gnome-keyring-daemon leaks memory

This patch from upstream bug #684351 seems to fix the problem here for me.

The attachment "patch to fix memory leak" of this bug report has been identified as being a patch. The ubuntu-reviewers team has been subscribed to the bug report so that they can review the patch. In the event that this is in fact not a patch you can resolve this situation by removing the tag 'patch' from the bug report and editing the attachment so that it is not flagged as a patch. Additionally, if you are member of the ubuntu-reviewers team please also unsubscribe the team from this bug report.

[This is an automated message performed by a Launchpad user owned by Brian Murray. Please contact him regarding any issues with the action taken in this bug report.]

Thanks Brian, do you have details on how to trigger the leak? It's not happening to every user and to reach that usage you probably have a client doing lot of keyring calls? I wonder if we could get a testcase and a improve the testsuit once the bug fixed to make sure the issue doesn't come back in such scenarios

Upstream wrote on bugzilla that "The leak is not cumulative" so it's a bit weird that it would lead to this situation, in any case since the fix seems right and works for you we can SRU it to precise, I would like to have a reply about some of those details first if possible though

On 13-01-07 01:06 PM, Sebastien Bacher wrote:
> Thanks Brian, do you have details on how to trigger the leak?

Nothing more specific than just "sit back and wait". :-) Seriously, I
don't to anything special here. I keep SSH and PGP keys in my keyring
and ssh a lot and sign every outgoing e-mail. Both ssh and e-mail
signing utilize the keyring of course.

The patch is applied in gnome-keyring >= 3.6.1, so marking as Fix Released. If you want to get this applied in previous releases, please follow instructions at <https://wiki.ubuntu.com/StableReleaseUpdates#Procedure>.

On 13-03-08 08:08 AM, Dmitry Shachnev wrote:
> The patch is applied in gnome-keyring >= 3.6.1,

Will that release go into the current LTS, 12.04? LTS is Long Term
Support so I would expect it to see bug fixes without requiring
individuals to file specific requests to have a bug fixed in it.

I don't have a 12.04 machine to build/test the fix on, maybe someone else can do that.

Or you can apply the patch yourself and submit a merge proposal against lp:ubuntu/precise/gnome-keyring, so that it'll show up in the sponsorship queue. See <http://developer.ubuntu.com/packaging/html/fixing-a-bug.html> for details.

But I do have a 12.04 machine and I already submitted a patch (which just came from upstream, not my patch) which fixes the problem on the gnome-keyring shipped in 12.04 LTS. So really, I don't know why you are proposing all of this rigmarole with backporting patches from g-k-r 3.6.1 and sponsorship, etc. Why not just ship the LTS 3.2.2 g-k-r release with the attached patch?

It's all of this beauraucratic red-tape to getting such a simple but important fix into what is supposed to be a supported release that makes Ubuntu no longer a desired platform to use.

> It's all of this beauraucratic red-tape to getting such a simple but important fix into what is supposed to be a supported release that makes Ubuntu no longer a desired platform to use.

oh come on, what "all of this beauraucratic red-tape", bugs just need basic infos on
- impact (what's the issue)
- test case (so the fix can be confirmed)
- regression potential (so testers know if they should watch for a potentiel regression)

it's basic infos that takes a few minutes to write (if you don't need to write a program as testcase)

then you need somebody to do the upload, anyone is welcome to help and provide a debdiff to help on that rather than complaining that volunteers don't work enough or not in a reasonable timeline fashion

@Brian: I will SRU that fix, I've marked it on my todolist for a while, but there is only so many hours in a week and only so many developers to maintain some thousands packages accross several release of Ubuntu, such comments don't make me feel motivated to do the work though :-(

uploaded to the precise review queue: http://launchpadlibrarian.net/133350726/gnome-keyring_3.2.2-2ubuntu4.1_source.changes

Hello Brian, or anyone else affected,

Accepted gnome-keyring into precise-proposed. The package will build now and be available at http://launchpad.net/ubuntu/+source/gnome-keyring/3.2.2-2ubuntu4.1 in a few hours, and then in the -proposed repository.

Please help us by testing this new package. See https://wiki.ubuntu.com/Testing/EnableProposed for documentation how to enable and use -proposed. Your feedback will aid us getting this update out to other Ubuntu users.

If this package fixes the bug for you, please add a comment to this bug, mentioning the version of the package you tested, and change the tag from verification-needed to verification-done. If it does not fix the bug for you, please add a comment stating that, and change the tag to verification-failed. In either case, details of your testing will help us make a better decision.

Further information regarding the verification process can be found at https://wiki.ubuntu.com/QATeam/PerformingSRUVerification . Thank you in advance!

Brian, can you confirm that the SRUed version fixes the problem for you? We need some confirmation of the bugfix before it can be published to precise-updates.

As Seb pointed out earlier, this problem is not reproducible for all users. By that measure, the test case in the bug description is incomplete, because it does not tell users how to reproduce the problem.

Dear Steve Langasek,

> Brian, can you confirm that the SRUed version fixes the problem for you?
> We need some confirmation of the bugfix before it can be published to
> precise-updates.
>
> As Seb pointed out earlier, this problem is not reproducible for all
> users. By that measure, the test case in the bug description is
> incomplete, because it does not tell users how to reproduce the problem.

I have installed the SRUed version at home (at least, I think I did it
correctly):

ii gnome-keyring 3.2.2-2ubuntu4 GNOME keyring services (daemon and tools)

It is running normally and not causing any problems at this moment.
Memory usage seems relatively normal:

hein 1899 0.0 0.0 368880 3732 ? Sl 21:01 0:00 /usr/bin/gnome-keyring-daemon --daemonize --login

although the virtual use does seem high but I see that with a lot of
programs on 64 bit ubuntu. I will try it out at work where I actually
noticed the memory leak and confirm again once I see it running a bit
longer.

Kind regards,
     Hein

--

 Hein Zelle <email address hidden>
                         http://www.icce.rug.nl/~hein

Hein, have you had a chance to test this at work?

Hello Hein,

Were you able to test this in your reproducable environment yet?

Thanks!
Adam

I've finally been able to test the new version on both my home machine and the work machine that originally showed the problem. Results seem positive: memory usage appears normal and doesn't grow after a full day of running.

directly after starting:

> zelle 24481 0.1 0.0 216800 4304 ? SLl 15:54 0:00 /usr/bin/gnome-keyring-daemon --daemonize --login

30 minutes and some use later:

> zelle 24481 0.0 0.0 364264 4432 ? SLl 15:54 0:00 /usr/bin/gnome-keyring-daemon --daemonize --login

The next day (approx. 24 hours):

> zelle 24481 0.0 0.0 437996 4104 ? SLl May01 0:00 /usr/bin/gnome-keyring-daemon --daemonize --login

There doesn't appear to be a memory leak, however the virtual memory allocation shown by ps is still significant with 440 Mb. I see this with lots of programs on 64 bit ubuntu, though, so that may be another problem altogether.

Note that I have disabled some of the xdg services in the mean time, which may prevent me from seeing the problem return. I've disabled the following services (XFCE, located in ~/.config/autostart):

> gnome-keyring-gpg.desktop
> gnome-keyring-secrets.desktop
> gnome-keyring-pkcs11.desktop
> gnome-keyring-ssh.desktop

Instead, I now only start gnome-keyring once with the following options:

> gnome-keyring-daemon --start --components=pkcs11,secrets,ssh,gpg

I'm not sure if this is relevant for the bug(fix), but with that last commandline everything seems to work fine now.

The verification of this Stable Release Update has completed successfully and the package has now been released to -updates. Subsequently, the Ubuntu Stable Release Updates Team is being unsubscribed and will not receive messages about this bug report. In the event that you encounter a regression using the package from -updates please report a new bug using ubuntu-bug and tag the bug report regression-update so we can easily find any regresssions.

This bug was fixed in the package gnome-keyring - 3.2.2-2ubuntu4.1

---------------
gnome-keyring (3.2.2-2ubuntu4.1) precise; urgency=low

  * debian/patches/git_leak_fix.patch: "rpc-layer: Fix memory leak in call
    state pool", thanks Brian J. Murrell for pointing it (lp: #1094496)
 -- Sebastien Bacher <email address hidden> Fri, 08 Mar 2013 17:04:41 +0100

Another way to reproduce that bug:

ssh-keygen -t rsa -N "" -f ~/.ssh/newkey
ssh-copy-id -i ~/.ssh/newkey.pub localhost
while true; do ssh localhost date; done

Then sit and observe in another terminal how memory disappears.

@wyderkat: which gnome-keyring version is that? There were *lots* of leak fixes in 3.16 release.

Tested with chrome making the login on X website with my credentials saved, gnome-keyring memory rose , logout and login on the same X website keyring memory rose again... by the end of the week gnome-keyring is using almost 200mb : O
(a lot of sites and ecomerce have timed logins so it happens a lot )

will the fix get to ubuntu 14.04 ?

UBUNTU 14.04.03 3.13.0-67-generic
google-chrome-stable 46.0.2490.80-1
gnome-keyring 3.10.1-1ubuntu4.2
Xfce 4.12

thanks.